Critical Care Department, "Attiko" University Hospital, Athens, Greece.
Am J Crit Care. 2011 Jan;20(1):56-66. doi: 10.4037/ajcc2011763.
Intensivists and nursing staff are often reluctant to admit patients with cancer to the intensive care unit even though these patients' survival rate has improved since the 1980s.
To identify factors associated with mortality in cancer patients admitted to the intensive care unit and to assess and compare the effectiveness of 3 general prognostic models: the Acute Physiology and Chronic Health Evaluation (APACHE) II, the Simplified Acute Physiology Score (SAPS II), and the Sequential Organ Failure Assessment (SOFA).
A prospective observational cohort study was performed in 2 general intensive care units. Discrimination was assessed by using area under the receiver operating characteristic curves, and calibration was evaluated by using Hosmer-Lemeshow goodness-of-fit tests.
A total of 126 patients were included during a 3-year period. The observed mortality was 46.8%. All 3 general models showed excellent discrimination (area under the curve >0.8) and good calibration (P = .17, .14, and .22 for APACHE II, SAPS II, and SOFA, respectively). However, discrimination was significantly better with APACHE II scores than with SOFA scores (P = .02). Multivariate analyses indicated that independent of the 3 severity-of-illness scores, unfavorable risk factors for mortality included a patient's preadmission performance status, source of admission (internal medicine vs surgery department), and the presence of septic shock, infection, or anemia. Combining SOFA and SAPS II scores with these variables created prognostic models with improved calibration and discrimination.
The general prognostic models seem fairly accurate in the prediction of mortality in critically ill cancer patients in the intensive care unit.
尽管自 20 世纪 80 年代以来癌症患者的存活率有所提高，但重症监护医师和护理人员往往不愿将癌症患者收入重症监护病房。
确定收入重症监护病房的癌症患者的死亡率相关因素，并评估和比较 3 种通用预后模型（急性生理学和慢性健康评估 [APACHE] II 评分、简化急性生理学评分 [SAPS] II 评分和序贯器官衰竭评估 [SOFA]）的有效性。
在 2 个普通重症监护病房进行了前瞻性观察性队列研究。通过使用接受者操作特征曲线下的面积来评估判别能力，并通过 Hosmer-Lemeshow 拟合优度检验来评估校准效果。
在 3 年期间共纳入 126 例患者。观察到的死亡率为 46.8%。所有 3 种通用模型均显示出优异的判别能力（曲线下面积>0.8）和良好的校准效果（APACHE II、SAPS II 和 SOFA 的 P 值分别为.17、.14 和.22）。然而，APACHE II 评分的判别能力明显优于 SOFA 评分（P=0.02）。多变量分析表明，独立于 3 种疾病严重程度评分，死亡率的独立危险因素包括患者入院前的身体状况、入院来源（内科与外科）以及脓毒性休克、感染或贫血的存在。将 SOFA 和 SAPS II 评分与这些变量相结合，创建了具有改善校准和判别能力的预后模型。
这些通用的预后模型似乎能够较为准确地预测重症监护病房中癌症危重症患者的死亡率。
